
#import libraries
import pandas  as pd 
import numpy as np 
import matplotlib.pyplot as plt 
import seaborn as sns
from sklearn.preprocessing import StandardScaler
from sklearn.model_selection import train_test_split
from sklearn import svm
from sklearn.metrics import accuracy_score
import warnings
warnings.filterwarnings('ignore')


#load the data
diabetes = pd.read_csv("diabetes.csv")
diabetes.head()


diabetes.shape

(15000, 10)


#show dataset information
diabetes.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 15000 entries, 0 to 14999
Data columns (total 10 columns):
 #   Column                  Non-Null Count  Dtype  
---  ------                  --------------  -----  
 0   PatientID               15000 non-null  int64  
 1   Pregnancies             15000 non-null  int64  
 2   PlasmaGlucose           15000 non-null  int64  
 3   DiastolicBloodPressure  15000 non-null  int64  
 4   TricepsThickness        15000 non-null  int64  
 5   SerumInsulin            15000 non-null  int64  
 6   BMI                     15000 non-null  float64
 7   DiabetesPedigree        15000 non-null  float64
 8   Age                     15000 non-null  int64  
 9   Diabetic                15000 non-null  int64  
dtypes: float64(2), int64(8)
memory usage: 1.1 MB


#CONVERT FLOAT TO INTEGERS
diabetes["BMI"] = diabetes["BMI"].astype(int)
diabetes["DiabetesPedigree"] = diabetes["DiabetesPedigree"].astype(int)
diabetes.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 15000 entries, 0 to 14999
Data columns (total 10 columns):
 #   Column                  Non-Null Count  Dtype
---  ------                  --------------  -----
 0   PatientID               15000 non-null  int64
 1   Pregnancies             15000 non-null  int64
 2   PlasmaGlucose           15000 non-null  int64
 3   DiastolicBloodPressure  15000 non-null  int64
 4   TricepsThickness        15000 non-null  int64
 5   SerumInsulin            15000 non-null  int64
 6   BMI                     15000 non-null  int32
 7   DiabetesPedigree        15000 non-null  int32
 8   Age                     15000 non-null  int64
 9   Diabetic                15000 non-null  int64
dtypes: int32(2), int64(8)
memory usage: 1.0 MB


#determine number with diabetes and those without
diabetes.Diabetic.value_counts()

0    10000
1     5000
Name: Diabetic, dtype: int64


#value count for pregnancies verses number of children
diabetes.Pregnancies.value_counts()

0     4377
1     2864
3     1219
6     1070
2      939
7      895
9      881
5      687
4      675
8      645
10     450
11     143
13      69
12      57
14      29
Name: Pregnancies, dtype: int64


##Check for missing value
plt.figure(figsize=(15,6))
sns.heatmap(diabetes.isnull(),cbar=False,cmap="viridis",yticklabels=False)
plt.title('Missing value in the dataset',fontsize=6)

Text(0.5, 1.0, 'Missing value in the dataset')


diabetes.describe()


#binning the age,BMI,PlasmaGlucose,Pregnancies,DiastolicBloodPressure,SerumInsulin data
diabetes["Age_group"] = pd.cut(diabetes['Age'],bins = [20,35,55,80], labels = ["young_adults","middle_age","old"])
diabetes["BMI_chart"] = pd.cut(diabetes['BMI'],bins = [0,18,25,30,35,40,np.inf], labels = ["LOW", "NORMAL", "OVERWEIGHT","OBESE 1","OBESE 2","0BESE 3"])
diabetes["PlasmaGlucose_cat"] = pd.cut(diabetes['PlasmaGlucose'], bins = [0,100,125,np.inf], labels = ["good","moderate","high"])
diabetes["Pregnancies_grp"] = pd.cut(diabetes['Pregnancies'], bins = [-1,5,10,15], labels = ["0-5","6-10","11-15"])
diabetes["DiastolicBloodPressure_chart"] = pd.cut(diabetes["DiastolicBloodPressure"], bins = [0,80,89,np.inf], labels = ["Normal","Elevated","Highbloodpressure"])
diabetes["SerumInsulin_chart"] = pd.cut(diabetes["SerumInsulin"], bins = [0,70,108,145,180,np.inf], labels = ["Lowbloodsugar","Optimal","Prediabetic","Diabetic","Deadly"])

diabetes.head()


#categorical data plus target
categorical_data=["Age_group","BMI_chart","PlasmaGlucose_cat","DiastolicBloodPressure_chart","SerumInsulin_chart","Pregnancies_grp"]
categorical_data=diabetes.drop(columns=["PatientID","Age","TricepsThickness","BMI","PlasmaGlucose","DiastolicBloodPressure","SerumInsulin","DiabetesPedigree","Pregnancies"])
categorical_data.head(8)


#continous data plus Diabetic as the target
diabetes_1=["Age","BMI","PlasmaGlucose","DiastolicBloodPressure","SerumInsulin","DiabetesPedigree","Pregnancies"]
diabetes_1=diabetes.drop(columns=["PatientID","Age_group","BMI_chart","PlasmaGlucose_cat","DiastolicBloodPressure_chart","SerumInsulin_chart","Pregnancies_grp"])
diabetes_1.head()


#A heatmap for continous data
plt.subplots(figsize=(35,25))
sns.set(font_scale=4.0)
sns.heatmap(diabetes_1.corr(),annot=True,cmap="viridis")

<AxesSubplot:>


#histogram visualization of continous data
import matplotlib.pyplot as plt
%matplotlib inline   
import matplotlib.pyplot as plt
diabetes_1.hist(figsize=(35,25))
plt.show()


### scatter plot the target verses continous data
plt.figure(figsize=(55,40))
sns.set(font_scale=1.0)
sns.pairplot(diabetes_1, hue ="Diabetic",palette="viridis_r")
plt.show()

<Figure size 3960x2880 with 0 Axes>


#data visualization of pregnancies vs Diabetic
plt.figure(figsize=(15,10))
sns.barplot(x="Diabetic", y="Pregnancies",hue="Pregnancies_grp",data=diabetes,palette="viridis_r")
plt.title("The relationship between Pregnancies vs Diabetic")

Text(0.5, 1.0, 'The relationship between Pregnancies vs Diabetic')


#data visualization of pregnancies vs Diabetic
plt.figure(figsize=(15,15))
sns.barplot(x="Diabetic", y="BMI",hue="BMI_chart",data=diabetes,palette="viridis_r")
plt.title("The relationship between the Diabetic Status and BMI Status")

Text(0.5, 1.0, 'The relationship between the Diabetic Status and BMI Status')


#data visualization of SerumInsulin vs Diabetic
plt.figure(figsize=(15,10))
sns.barplot(x="Diabetic", y="SerumInsulin",hue="SerumInsulin_chart",data=diabetes,palette="viridis_r")
plt.title("Bar plot of SerumInsulin vs Diabetic")

Text(0.5, 1.0, 'Bar plot of SerumInsulin vs Diabetic')


#data visualization of SerumInsulin vs Diabetic
plt.figure(figsize=(15,10))
sns.violinplot(x="Diabetic", y="PlasmaGlucose",hue="PlasmaGlucose_cat",data=diabetes,palette="viridis_r")
plt.title("violin plot of PlasmaGlucose vs Diabetic attributes")

Text(0.5, 1.0, 'violin plot of PlasmaGlucose vs Diabetic attributes')


#data visualization of Age vs Diabetic
plt.figure(figsize=(15,10))
sns.barplot(x="Diabetic", y="Age",hue="Age_group",data=diabetes,palette="viridis_r")
plt.title("Bar plot of Age vs Diabetic attributes")

Text(0.5, 1.0, 'Bar plot of Age vs Diabetic attributes')


#data visualization of Age vs Diabetic
plt.figure(figsize=(15,12))
sns.violinplot(x="Diabetic", y="DiastolicBloodPressure",hue="DiastolicBloodPressure_chart",data=diabetes,palette="viridis_r")
plt.title("Bar plot of Age vs Diabetic attributes")

Text(0.5, 1.0, 'Bar plot of Age vs Diabetic attributes')


f, ax = plt.subplots(figsize=(10, 10))
ax = sns.barplot(x="Diabetic", y="DiabetesPedigree", data=diabetes_1,palette="viridis_r")
plt.show()


fig, axes = plt.subplots(2,3, figsize=(20, 15))

fig.suptitle('')

sns.boxplot(ax=axes[0, 0], data=diabetes, x="Diabetic", y='Pregnancies',palette="viridis_r")
sns.boxplot(ax=axes[0, 1], data=diabetes, x="Diabetic", y='Age',palette="viridis_r")
sns.boxplot(ax=axes[0, 2], data=diabetes, x="Diabetic", y='PlasmaGlucose',palette="viridis_r")
sns.boxplot(ax=axes[1, 0], data=diabetes, x="Diabetic", y="BMI",palette="viridis_r")
sns.boxplot(ax=axes[1, 1], data=diabetes, x="Diabetic", y="DiastolicBloodPressure",palette="viridis_r")
sns.boxplot(ax=axes[1, 2], data=diabetes, x="Diabetic", y="DiabetesPedigree",palette="viridis_r")

<AxesSubplot:xlabel='Diabetic', ylabel='DiabetesPedigree'>


#determine the interquantile range
Q1 = diabetes_1.quantile(0.25)
Q3 = diabetes_1.quantile(0.75)
IQR = Q3 - Q1
IQR

Pregnancies                 6.0
PlasmaGlucose              45.0
DiastolicBloodPressure     27.0
TricepsThickness           26.0
SerumInsulin              156.0
BMI                        18.0
DiabetesPedigree            0.0
Age                        13.0
Diabetic                    1.0
dtype: float64


(diabetes_1 < (Q1 - 1.5 * IQR)) |(diabetes_1> (Q3 + 1.5 * IQR))


diabetes_out = diabetes_1[~((diabetes_1< (Q1 - 1.5 * IQR)) |(diabetes_1 > (Q3 + 1.5 * IQR))).any(axis=1)]
diabetes_out.shape

(12358, 9)


diabetes_1=diabetes_out
diabetes_1.head()


diabetes_1.info()

<class 'pandas.core.frame.DataFrame'>
Int64Index: 12358 entries, 1 to 14998
Data columns (total 9 columns):
 #   Column                  Non-Null Count  Dtype
---  ------                  --------------  -----
 0   Pregnancies             12358 non-null  int64
 1   PlasmaGlucose           12358 non-null  int64
 2   DiastolicBloodPressure  12358 non-null  int64
 3   TricepsThickness        12358 non-null  int64
 4   SerumInsulin            12358 non-null  int64
 5   BMI                     12358 non-null  int32
 6   DiabetesPedigree        12358 non-null  int32
 7   Age                     12358 non-null  int64
 8   Diabetic                12358 non-null  int64
dtypes: int32(2), int64(7)
memory usage: 868.9 KB


fig, axes = plt.subplots(2,3, figsize=(20, 15))

fig.suptitle('')

sns.boxplot(ax=axes[0, 0], data=diabetes_1, x="Diabetic", y='Pregnancies',palette="viridis_r")
sns.boxplot(ax=axes[0, 1], data=diabetes_1, x="Diabetic", y='Age',palette="viridis_r")
sns.boxplot(ax=axes[0, 2], data=diabetes_1, x="Diabetic", y='PlasmaGlucose',palette="viridis_r")
sns.boxplot(ax=axes[1, 0], data=diabetes_1, x="Diabetic", y="BMI",palette="viridis_r")
sns.boxplot(ax=axes[1, 1], data=diabetes_1, x="Diabetic", y="DiastolicBloodPressure",palette="viridis_r")
sns.boxplot(ax=axes[1, 2], data=diabetes_1, x="Diabetic", y="DiabetesPedigree",palette="viridis_r")

<AxesSubplot:xlabel='Diabetic', ylabel='DiabetesPedigree'>


#Create X (all the feature columns)
X= diabetes_1.drop("Diabetic", axis=1)#Features

# Create y (the target column)
Y= diabetes_1["Diabetic"]#labels
X.head()


Y.head()

1    0
2    0
4    0
5    0
6    0
Name: Diabetic, dtype: int64


diabetes_1.shape

(12358, 9)


scaler=StandardScaler()


scaler.fit(X)

StandardScaler()


standardized_data=scaler.transform(X)
print(standardized_data)

[[ 1.46969417 -0.48249464  1.29989545 ... -0.99021237  0.
  -0.53215359]
 [ 1.17247481  0.23516611 -1.4189578  ...  1.03733252  0.
  -0.53215359]
 [-0.61084134 -0.70091313 -0.70969173 ...  1.13870977  0.
  -0.64646731]
 ...
 [-0.31362198 -1.07534483 -0.29595319 ... -0.07781717  0.
   1.18255222]
 [-0.9080607  -0.451292    1.06347343 ... -1.2943441   0.
  -0.41783987]
 [-0.9080607   0.76561102  1.59542298 ... -1.19296686  0.
  -0.53215359]]


scaler.mean_

array([  3.05518692, 107.46326266,  71.00720181,  28.45905486,
       115.67624211,  30.76759994,   0.        ,  27.65520311])


scaler.scale_

array([ 3.36451839, 32.04856883, 16.91889767, 14.27595685, 96.20242592,
        9.86414657,  1.        ,  8.74785628])


#splitting the training and test data
from sklearn.model_selection import train_test_split
data = diabetes_1
X_train,X_test,Y_train,Y_test = train_test_split(X,Y,test_size=0.3,stratify=Y,random_state=42)

X_train.shape,X_test.shape, Y_train.shape, Y_test.shape

((8650, 8), (3708, 8), (8650,), (3708,))


from sklearn.linear_model import LogisticRegression
from sklearn.neighbors import KNeighborsClassifier
from sklearn.svm import SVC
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import RandomForestClassifier


classifiers = {
    "LogisticRegression" : LogisticRegression(),
    "KNeighbors" : KNeighborsClassifier(),
    "DecisionTree" : DecisionTreeClassifier(),
    "RandomForest" : RandomForestClassifier()}


from sklearn.linear_model import LogisticRegression
import sklearn.metrics as metrics
regression = LogisticRegression(max_iter = 1000)
regression.fit(X_train, Y_train )
y_pred =regression.predict(X_test)
print("Accuracy Score:",metrics.accuracy_score(Y_test, y_pred))
print("Train Accuracy:",regression.score(X_train,Y_train))
print("Test Accuracy:",regression.score(X_test,Y_test))

Accuracy Score: 0.8071736785329018
Train Accuracy: 0.8006936416184971
Test Accuracy: 0.8071736785329018


from sklearn import tree
from sklearn.metrics import accuracy_score
decision_tree = tree.DecisionTreeClassifier()
decision_tree = decision_tree.fit(X_train,Y_train)
y_pred = decision_tree.predict(X_test)
print("Accuracy Score:",metrics.accuracy_score(Y_test, y_pred))
print("Train Accuracy:",decision_tree.score(X_train,Y_train))
print("Test Accuracy:",decision_tree.score(X_test,Y_test))

Accuracy Score: 0.901564185544768
Train Accuracy: 1.0
Test Accuracy: 0.901564185544768


from sklearn.ensemble import RandomForestClassifier
from sklearn import metrics
#Create a Gaussian Classifier
clf=RandomForestClassifier(n_estimators=1000)
#Train the model using the training sets y_pred=clf.predict(X_test)
clf.fit(X_train,Y_train)
y_pred=clf.predict(X_test)
print("Accuracy Score:",metrics.accuracy_score(Y_test, y_pred))
print("Train Accuracy:",clf.score(X_train,Y_train))
print("Test Accuracy:",clf.score(X_test,Y_test))

Accuracy Score: 0.9390507011866235
Train Accuracy: 1.0
Test Accuracy: 0.9390507011866235


from sklearn.neighbors import KNeighborsClassifier# Create KNN classifier
knn = KNeighborsClassifier(n_neighbors = 3)# Fit the classifier to the data
knn.fit(X_train,Y_train)
y_pred = knn.predict(X_test)
print("Accuracy Score:",metrics.accuracy_score(Y_test, y_pred))
print("Train Accuracy:",knn.score(X_train,Y_train))
print("Test Accuracy:",knn.score(X_test,Y_test))

Accuracy Score: 0.8460086299892126
Train Accuracy: 0.9186127167630058
Test Accuracy: 0.8460086299892126


from sklearn.metrics import classification_report, confusion_matrix
from sklearn.metrics import ConfusionMatrixDisplay
y_pred = clf.predict(X_test)
cm = confusion_matrix(Y_test,y_pred, labels=clf.classes_)
print(confusion_matrix(Y_test,y_pred))
print(classification_report(Y_test,y_pred))
disp = ConfusionMatrixDisplay(confusion_matrix=cm,display_labels=clf.classes_)

# labels, title and ticks
ax.set_xlabel('Predicted labels');ax.set_ylabel('True labels'); 
ax.set_title('Confusion Matrix')

disp.plot()

[[2620   68]
 [ 158  862]]
              precision    recall  f1-score   support

           0       0.94      0.97      0.96      2688
           1       0.93      0.85      0.88      1020

    accuracy                           0.94      3708
   macro avg       0.94      0.91      0.92      3708
weighted avg       0.94      0.94      0.94      3708

<sklearn.metrics._plot.confusion_matrix.ConfusionMatrixDisplay at 0x1d8a1127ee0>


from sklearn import metrics
import numpy as np
import matplotlib.pyplot as plt
from sklearn.metrics import roc_curve
y_pred = clf.predict(X_test)
clf_fpr,clf_tpr, threshold = roc_curve(Y_test, y_pred)

plt.figure(figsize=(20,15), dpi=80)
plt.plot(clf_fpr, clf_tpr, linestyle='-',)
plt.xlabel('False Positive Rate -->')
plt.ylabel('True Positive Rate -->')
plt.title("ROC curve for diabetes classifier (clf)")
plt.legend()

plt.show()

No handles with labels found to put in legend.


clf.feature_importances_

array([0.29744027, 0.10277835, 0.06333773, 0.07150852, 0.14396955,
       0.16204751, 0.        , 0.15891807])


# random forest for feature importance on a classification problem
from sklearn.datasets import make_classification
from sklearn.ensemble import RandomForestClassifier
from matplotlib import pyplot
# define dataset
features,label = make_classification(n_samples=1000, n_features=8, n_informative=3, n_redundant=3, random_state=1)
# define the model
model = RandomForestClassifier()
# fit the model
model.fit(features,label)
# get importance
importance = model.feature_importances_
# summarize feature importance
for i,vi in enumerate(importance):
    print('Features: %0d, Score: %.5f' % (i,vi))
# plot feature importance
pyplot.bar([x for x in range(len(importance))], importance)
pyplot.show()

Features: 0, Score: 0.08031
Features: 1, Score: 0.14574
Features: 2, Score: 0.01712
Features: 3, Score: 0.02024
Features: 4, Score: 0.09528
Features: 5, Score: 0.09094
Features: 6, Score: 0.38908
Features: 7, Score: 0.16129


import pickle
# Save trained model to file
y_pred = clf.predict(X_test)
pickle.dump(clf, open("Diabetes.pkl", "wb"))
loaded_model = pickle.load(open("Diabetes.pkl", "rb"))
loaded_model.predict(X_test)
loaded_model.score(X_test,Y_test)
clf.predict(X_test)

array([0, 1, 0, ..., 0, 0, 0], dtype=int64)


from pandas_profiling import ProfileReport


profile = ProfileReport(diabetes)
profile

	PatientID	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	DiabetesPedigree	Age	Diabetic
count	1.500000e+04	15000.000000	15000.000000	15000.000000	15000.000000	15000.000000	15000.000000	15000.000000	15000.000000	15000.000000
mean	1.502922e+06	3.224533	107.856867	71.220667	28.814000	137.852133	31.007733	0.080867	30.137733	0.333333
std	2.892534e+05	3.391020	31.981975	16.758716	14.555716	133.068252	9.763696	0.286472	12.089703	0.471420
min	1.000038e+06	0.000000	44.000000	24.000000	7.000000	14.000000	18.000000	0.000000	21.000000	0.000000
25%	1.252866e+06	0.000000	84.000000	58.000000	15.000000	39.000000	21.000000	0.000000	22.000000	0.000000
50%	1.505508e+06	2.000000	104.000000	72.000000	31.000000	83.000000	31.000000	0.000000	24.000000	0.000000
75%	1.755205e+06	6.000000	129.000000	85.000000	41.000000	195.000000	39.000000	0.000000	35.000000	1.000000
max	1.999997e+06	14.000000	192.000000	117.000000	93.000000	799.000000	56.000000	2.000000	77.000000	1.000000

1.LOAD THE DATA¶

CONVERT FLOAT TO INTEGERS¶

COUNT¶

2.Exploratory data analysis¶

BINNING THE DATA¶

CATEGORICAL DATA¶

CONTINOUS DATA¶

DATA VISULIZATION¶

HANDLING OUTLIERS¶

DETECTING OUTLIERS¶

REMOVE OUTLIERS¶

SEPARATING FEATURES FROM THE LABEL¶

FEATURE SCALING¶

SPLITTING THE DATA¶

3. MODEL BUILDING¶

Logistic Regression¶

Decision Tree¶

Random Forest Classifier¶

K-Nearest Neighbors¶

RandomForestClassifier model,is the best of the 4,as it has the highest Accuracy percentage score¶

4. MODEL PERFOMANCE ANALYSIS¶

CLASSIFICATION REPORT AND CONFUSION MATRIX¶

ROC_CURVE¶

FEATURE IMPORTANCE¶

PANDA PROFILING REPORT¶

	PatientID	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	DiabetesPedigree	Age	Diabetic
0	1354778	0	171	80	34	23	43.509726	1.213191	21	0
1	1147438	8	92	93	47	36	21.240576	0.158365	23	0
2	1640031	7	115	47	52	35	41.511523	0.079019	23	0
3	1883350	9	103	78	25	304	29.582192	1.282870	43	1
4	1424119	1	85	59	27	35	42.604536	0.549542	22	0

	PatientID	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	DiabetesPedigree	Age	Diabetic	Age_group	BMI_chart	PlasmaGlucose_cat	Pregnancies_grp	DiastolicBloodPressure_chart	SerumInsulin_chart
0	1354778	0	171	80	34	23	43	1	21	0	young_adults	0BESE 3	high	0-5	Normal	Lowbloodsugar
1	1147438	8	92	93	47	36	21	0	23	0	young_adults	NORMAL	good	6-10	Highbloodpressure	Lowbloodsugar
2	1640031	7	115	47	52	35	41	0	23	0	young_adults	0BESE 3	moderate	6-10	Normal	Lowbloodsugar
3	1883350	9	103	78	25	304	29	1	43	1	middle_age	OVERWEIGHT	moderate	6-10	Normal	Deadly
4	1424119	1	85	59	27	35	42	0	22	0	young_adults	0BESE 3	good	0-5	Normal	Lowbloodsugar

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	DiabetesPedigree	Age	Diabetic
0	0	171	80	34	23	43	1	21	0
1	8	92	93	47	36	21	0	23	0
2	7	115	47	52	35	41	0	23	0
3	9	103	78	25	304	29	1	43	1
4	1	85	59	27	35	42	0	22	0

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	DiabetesPedigree	Age	Diabetic
0	False	False	False	False	False	False	True	False	False
1	False	False	False	False	False	False	False	False	False
2	False	False	False	False	False	False	False	False	False
3	False	False	False	False	False	False	True	False	False
4	False	False	False	False	False	False	False	False	False
...	...	...	...	...	...	...	...	...	...
14995	False	False	False	False	False	False	False	False	False
14996	False	False	False	False	False	False	False	False	False
14997	False	False	False	False	False	False	False	False	False
14998	False	False	False	False	False	False	False	False	False
14999	False	False	False	False	True	False	False	False	False

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	Age
1	8	92	93	47	36	21	23
2	7	115	47	52	35	41	23
4	1	85	59	27	35	42	22
5	0	82	92	9	253	19	26
6	0	133	47	19	227	21	21

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	Age
1	8	92	93	47	36	21	23
2	7	115	47	52	35	41	23
4	1	85	59	27	35	42	22
5	0	82	92	9	253	19	26
6	0	133	47	19	227	21	21

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	DiabetesPedigree	Age	Diabetic
0	0	171	80	34	23	43	1	21	0
1	8	92	93	47	36	21	0	23	0
2	7	115	47	52	35	41	0	23	0
3	9	103	78	25	304	29	1	43	1
4	1	85	59	27	35	42	0	22	0

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	Age
1	8	92	93	47	36	21	23
2	7	115	47	52	35	41	23
4	1	85	59	27	35	42	22
5	0	82	92	9	253	19	26
6	0	133	47	19	227	21	21

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	Age
1	8	92	93	47	36	21	23
2	7	115	47	52	35	41	23
4	1	85	59	27	35	42	22
5	0	82	92	9	253	19	26
6	0	133	47	19	227	21	21

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	DiabetesPedigree	Age	Diabetic
0	0	171	80	34	23	43	1	21	0
1	8	92	93	47	36	21	0	23	0
2	7	115	47	52	35	41	0	23	0
3	9	103	78	25	304	29	1	43	1
4	1	85	59	27	35	42	0	22	0

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	Age
1	8	92	93	47	36	21	23
2	7	115	47	52	35	41	23
4	1	85	59	27	35	42	22
5	0	82	92	9	253	19	26
6	0	133	47	19	227	21	21

	Pregnancies	PlasmaGlucose	DiastolicBloodPressure	TricepsThickness	SerumInsulin	BMI	Age
1	8	92	93	47	36	21	23
2	7	115	47	52	35	41	23
4	1	85	59	27	35	42	22
5	0	82	92	9	253	19	26
6	0	133	47	19	227	21	21